Indian Palaeography c16.s1.12

Language: English

c16.s1.12.1

A. Pl. XIX. 1 and 2. The earliest examples of Ceylonese writing come from the caves of Vessagiri, Ritigala, Keramba Hina and Naulpata. These cave inscriptions are related to one another not only by their contents but also by the particular kind of the rock surface on which they are inscribed. Morphologically they are reducible into two groups. Inscriptions included under no. 1³ have long verticals as in the earliest inscriptions from India, but cave inscription no. 12 from Vessagiri, which falls under no. 2, has its verticals equalized—a practice which started in North India after the Mathura Kshatrapa inscriptions in the first century A.D.¹ But this practice did not reach Ceylon directly from the north, as is clear from the letters which are akin to southern characters. Hence no. 2 can hardly be dated earlier than the beginning of the second century A.D. Here we find only three medial vowels, *i*, *u* and *e*. Of these the medial *u*, when added to *pu* and *su*, is a downward extension of the right vertical—a later practice. The letter *ka* has its vertical longer than its cross bar. *Ta* has a hook attached to the right of a vertical in the Deccani style. The angularity is well marked in the forms of *pa*, *sa* and *ha*. *Va* has a triangular base. The letter *ma* copies the form of the southern caves. The conjunct *sra* has its *ra* appended at the right bottom of *sa*. There is a certain unity in the inscriptions included under no.

c16.s1.12.2

1, though here also different traditions are traceable. Of the initial vowels the most distinctive is *a*, where the two left curves meet the right vertical separately—a practice which became common in the post-Aśokan period.² The initial *u* generally has its vertical longer than the horizontal. The medial *i*, as in *ki*, *chi*, *di* and *ri*, starts from the middle of the vertical; the medial *u* is applied to the left, middle, or right of the base line, and the medial *o* in *śo* is a horizontal line drawn across the angular peak of *śa*. These features again point to a late date. In the letter forms two traditions are marked—one with rounded features, and the other with angular forms, especially in Ritigala cave inscriptions nos. 2–7. The angular features can be seen in the forms of *a*, *ḍa*, *da*, *pa*, the triangular base of *ma*, *la*, *sa* and *ha*. Occasionally the vertical of *ka* is equal in length to its cross-bar. *Ga* has an angular peak. *Gha* in Ritigala nos. 8–10 has its left vertical slightly longer than the remaining two. *Ṭha* has a flat base. Both *ṇa* and *na* have short bases. *Ta* is of the Deccani type with an inclined stroke to the right of a vertical. The mouth of *da* is open to the left. *Ba* is a square. *Bha* in Ritigala cave inscriptions nos. 8–10 follows the older form, but its upper stem is added by mistake to the left instead of to the right.

c16.s1.12.3

*Ma* has varying forms—the first is of the southern cave type with a horizontal line within bent arms, which has been called the Dravidian *ma*;³ the second is of the Nanaghat or Girnar style with two upper diagonals projecting from a circular base;¹ the third variety has its upper strokes curved; and the fourth variety, which occurs only once, has the upper curving strokes separated from the circular base by a line. *Ya* has a crescentic base with the middle vertical extended. *Ra* is either a straight or serpentine vertical line. *Va* usually has a rounded base. *Śa* has an angular peak, but its right limb is shorter than its left, and the mid-stroke is attached to the left limb. The writing in these inscriptions is clearly related to that of the Nanaghat records in Western India and that seen in the caves of the extreme south, though local tendencies in the application of the medial *i* and *o* and the peculiar forms of *śa*, *bha* and *ra* are obvious. In any case there is no reason to date these inscriptions earlier than their Indian counterparts, and according to the chronology of the Indian inscriptions already discussed² these inscriptions may also be dated about the beginning of the first century A.D.